JuliaRun is the twin of JuliaPro. While JuliaPro is meant for the developer or the data scientist who uses Julia for their own work, JuliaRun is meant to rapidly deploy and scale Julia applications on public and private cloud infrastructure. It takes the guesswork out of building scalable solutions and can be used by data scientists and engineers with little to no knowledge of how such systems need to be architected. Imagine developing an application using JuliaPro and deploying it in a scalable robust deployment environment with a single click.

Julia Computing has had the experience of running Julia programs in production since January 2015 on JuliaBox in a 24x7 environment serving more than 36,000 users from time-to-time. Based on the experience gained, we have extracted the deployment parts of JuliaBox into its own product JuliaRun, so that other users of Julia can also benefit from the architecture that we have put together.

JuliaRun architecture is designed for two kinds of workloads:

• Batch workloads: Jobs are submitted into a queue (which may originate from anywhere) and executed on the cluster.
• Real-time workloads: A Julia function is used to serve up a REST API endpoint, with JuliaRun managing the monitoring of arrivals, time to serve requests, and auto-scaling of containers.

JuliaRun supports two types of cluster managers:

• Static: Where a set of dedicated nodes are provided to the JuliaRun application.
• Elastic: Where nodes are dynamically allocated from a pool of available resources, either in a public or private cloud.

Currently, JuliaRun supports AWS and Google Cloud (and others can be integrated based on requirements).

Contact us at [email protected] for pricing options.